43483 - Copying text hangs browser exporting to XIF

Status: VERIFIED WORKSFORME

(Core :: DOM: Editor, defect, P3)

Description

[Mike Pinkerton (not reading bugmail)]

- got to www.shugashack.com
- select the text of the first news item "As a reminder, Stomped will be 
hosting..."
- cmd-c

the browser hangs trying to export the data to XIF in the selection code. Not 
sure if it would ever complete if i let it go for a long time.

happens on mac and win32

Comment 1

[Mike Pinkerton (not reading bugmail)]

cc'ing friend that told me about the bug, marking for beta2. I wonder how many 
other pages exhibit this behavior.

Comment 2

[rubydoo123]

assigning to Harish -- can you take a look at this one?

using the current win32 build:
1. this document contains 4 HTML start tags
2. the document has multiple close IFRAME and no open IFRAME
3. it contains microsoft unique web bots
4. the page takes an extremely long time to load, in the comm and debug builds
5. making a local copy ends with the same hang
6. I broke the file into the 4 different html files, document #1 and document #2 
are where the bad stuff resides
7. you don't even have to select anything to make it hang, just click

I will attach the 2 offending document segments and will continue to try and 
figure out what is making the app hang. I suspect either the web bots, iframes 
or js.

Comment 3

[rubydoo123]

Attachment: this is document #1

Comment 4

[rubydoo123]

Attachment: this is document #2

Comment 5

[rubydoo123]

in document 1, I did the following and it now works just fine:

1. added comment at beginning and end of style
2. moved the illegal script element within body to the legal location within the 
head
3. removed the hanging close TD within the 1st table
4. added the required space between the comment sto and web bot
5. removed the hanging end anchor tag in the 2nd table
6. removed the hanging end iframe tag in the 2nd table
7. added the required type attribute in the script element start start

The HTML was pretty messed up, out of all of the items that I repaired, the 
biggest offender was probably the illegal script element within the body and the 
missing required attribute. The hanging close elements were bad, but probably 
were not the major offenders. The malformed comment could also have been a 
contributor.

Comment 6

[rubydoo123]

In document #2, there were numerous issues:

1. two illegal script elements within the body
2. two body element start tags
3. a multitude of hanging end tags that I removed
4. reversed end tag sequences
5. several missing end tags
6. repaired an ending font tag, it was </font...

I gave up about a third of the way through the document. However, just doing 
what I did resulted in the page loading correctly and I am able to select and 
copy the text.

Comment 7

[rubydoo123]

Harish, it seems that if an illegal script element within the body, causes all 
sorts of problems with the document.

Comment 8

[Jay Patel [:jay]]

Putting on [nsbeta2-] radar. Not critical to beta2. 

Comment 9

[harishd]

Nominating for nsbeta3.

Comment 10

[harishd]

Here is the trace..where I see the hang:

nsDST::SearchTree(void * 0x03ea108c, unsigned int 0, void * * 0x0012f120) line 
563 + 6 bytes
nsDST::Search(void * 0x03ea108c, unsigned int 0, void * * 0x0012f120) line 331 + 
20 bytes
FrameManager::GetPrimaryFrameFor(FrameManager * const 0x03d15f00, nsIContent * 
0x03ea108c, nsIFrame * * 0x0012f120) line 488
PresShell::GetPrimaryFrameFor(const PresShell * const 0x03d167e0, nsIContent * 
0x03ea108c, nsIFrame * * 0x0012f120) line 3472 + 32 bytes
PresShell::GetGeneratedContentIterator(const PresShell * const 0x03d167e0, 
nsIContent * 0x03ea108c, nsIPresShell::GeneratedContentType Before, 
nsIContentIterator * * 0x047e2cd8) line 3070
nsGeneratedContentIterator::GetDeepFirstChild(nsCOMPtr<nsIContent> {...}) line 
446 + 67 bytes
nsGeneratedContentIterator::NextNode(nsCOMPtr<nsIContent> * 0x047e2cc0) line 682 
+ 24 bytes
nsGeneratedContentIterator::Next(nsGeneratedContentIterator * const 0x047e2cb0) 
line 788 + 15 bytes
nsDOMSelection::selectFrames(nsDOMSelection * const 0x03d16670, nsIPresContext * 
0x0417d100, nsIContentIterator * 0x047e2cb0, nsIContent * 0x03e89328, 
nsIDOMRange * 0x047e0290, nsIPresShell * 0x03d167e0, int 1) line 3495 + 12 bytes
nsDOMSelection::selectFrames(nsDOMSelection * const 0x03d16670, nsIPresContext * 
0x0417d100, nsIDOMRange * 0x047e0290, int 1) line 3568
nsDOMSelection::Extend(nsDOMSelection * const 0x03d16670, nsIDOMNode * 
0x03eb7a10, int 23) line 4730
nsSelection::TakeFocus(nsSelection * const 0x03d166f0, nsIContent * 0x03eb7a18, 
unsigned int 23, unsigned int 23, int 1, int 0) line 1720
nsSelection::HandleClick(nsSelection * const 0x03d166f0, nsIContent * 
0x03eb7a18, unsigned int 23, unsigned int 23, int 1, int 0, int 0) line 1585 + 
35 bytes
nsSelection::HandleDrag(nsSelection * const 0x03d166f0, nsIPresContext * 
0x0417d100, nsIFrame * 0x042968a4, nsPoint & {...}) line 1627 + 37 bytes
nsFrame::HandleDrag(nsFrame * const 0x042968a4, nsIPresContext * 0x0417d100, 
nsGUIEvent * 0x0012f8a8, nsEventStatus * 0x0012f798) line 1278
nsFrame::HandleEvent(nsFrame * const 0x042968a4, nsIPresContext * 0x0417d100, 
nsGUIEvent * 0x0012f8a8, nsEventStatus * 0x0012f798) line 787 + 27 bytes
PresShell::HandleEventInternal(nsEvent * 0x0012f8a8, nsIView * 0x03e45c10, 
nsEventStatus * 0x0012f798) line 3917 + 38 bytes
PresShell::HandleEven

Comment 11

[harishd]

Mike, could this be your area? Reassigning bug to [e-mail] and CCing 
myself.

Comment 12

[mjudge]

this is really making a whacky tree i will see if i can really discect this with 
joe's help.

Comment 13

[mjudge]

*** Bug 45650 has been marked as a duplicate of this bug. ***

Comment 14

[rubydoo123]

this site is such a mess that it really is one of thise edge cases, marking this 
as future. The engineer responsible for this area is doomed, if this impacts you 
or if this is imperative to your daily work, please add a comment and we will 
review the situation.

Comment 15

[anthonyd]

well, since the site is a mess, and also since XIF is gone, makring htis as 
WORKSFORME.

anthonyd

Comment 16

[sujay]

verified in 3/22 build.